A computer-implemented method for controlling an unmanned aerial vehicle (UAV) includes obtaining a first image captured by an imaging device carried by the UAV during a takeoff of the UAV from a target location, obtaining a second image from the imaging device in response to an indication to return to the target location, determining a spatial relationship between the UAV and the target location by comparing the first image and the second image, and controlling the UAV to approach the target location based at least in part on the spatial relationship.
METHODS AND SYSTEM FOR VISION-BASED LANDING
2019-07-18
Patent
Elektronische Ressource
Englisch
Multiple-Hypothesis Vision-Based Landing Autonomy
AIAA | 2020
|Falsification of a Vision-based Automatic Landing System
AIAA | 2021
|